[AMD] Support a new flag to disable quant on parallelLinear layer if required (#11811)

2025-10-23 10:16:15 +08:00
parent 6ade6a02d4
commit 99c92ff24b
2 changed files with 9 additions and 1 deletions
--- a/python/sglang/srt/environ.py
+++ b/python/sglang/srt/environ.py
@@ -158,6 +158,7 @@ class Envs:
    # AMD & ROCm
    SGLANG_USE_AITER = EnvBool(False)
    SGLANG_ROCM_FUSED_DECODE_MLA = EnvBool(False)
+    SGLANG_ROCM_DISABLE_LINEARQUANT = EnvBool(False)

    # Quantization
    SGLANG_INT4_WEIGHT = EnvBool(False)